The four-fifths rule is a rule of thumb from the US Uniform Guidelines on Employee Selection Procedures. Calculate the selection rate for each group, which is the number selected divided by the number who applied, then divide each group's rate by the highest group's rate. If the result is below 0.80, the guidelines say federal enforcement agencies will generally regard it as evidence of adverse impact. It is a quick check, not a verdict. With small numbers it can mislead, and passing does not prove a process is fair. Use it at each stage, including automated resume screening, as a prompt to look closer at why the rates differ and whether the selection step is job-related.
This is general information, not legal advice. Status as of October 2026.
The formula
- Selection rate for a group = number selected / number who applied (or entered the stage).
- Impact ratio for a group = its selection rate / the highest selection rate among groups.
- If the impact ratio is less than 0.80, adverse impact is indicated.
The text of the rule is in the Uniform Guidelines, 29 CFR 1607.4.
A worked example
Suppose 200 applicants pass through a resume screening stage.
| Group | Applicants | Passed | Selection rate | Impact ratio |
|---|---|---|---|---|
| Group A | 120 | 48 | 40% | 1.00 (highest) |
| Group B | 80 | 24 | 30% | 0.75 |
Group B's rate is 30 percent, which is 75 percent of Group A's 40 percent. That is below four-fifths, so the result flags possible adverse impact at this stage. The numbers are invented, but they show the method.
What the rule does not tell you
- Small samples: with a handful of applicants, one person changes the ratio a lot. Statisticians use significance tests, such as Fisher's exact test, alongside the ratio.
- Causes: a low ratio does not say why. It could be the criteria, the data, the way a question was worded or who applied.
- Legality: the guidelines treat it as evidence, not a final finding. An employer can justify a selection step that is job-related and consistent with business necessity, though it should also consider less discriminatory alternatives.
- Other measures: agencies and courts may use other statistics, particularly when samples are large.
- Other places: rules differ outside the US. Many countries apply similar ideas under different tests.
Where to apply it
Check each stage where people are filtered:
| Stage | Question |
|---|---|
| Application form and knock-outs | Do some groups fail the knock-outs more often? |
| Resume screening, manual or automated | Do some groups pass at lower rates? See AI bias in hiring |
| Tests and assessments | Do some groups score lower on a test? See skills assessment tests |
| Interviews | Do ratings differ by group for the same evidence? |
| Offers | Are offers made at different rates? |
If the overall process shows no adverse impact, the guidelines say agencies will usually not expect a review of each component. If it does, look at each stage.
Collecting the data
You need group information to calculate it. That raises its own issues.
- Collect demographic data only where lawful, voluntarily, with a clear explanation, and keep it separate from the decision makers.
- Store it securely and for no longer than needed. See candidate data retention.
- Where you cannot collect, use a vendor or an auditor who can, or analyse by proxy only with advice.
In the US, the guidelines expect users to keep records that show the impact of selection procedures by race, sex and ethnic group.
What to do if the ratio is low
- Check the data and the sample size.
- Look at the criteria. Is each tied to the job? See screening criteria scorecard.
- Check for proxies, such as location, school or gaps. See AI bias in hiring.
- Review the cutoff. A fixed score chosen without review can drive the gap. See resume ranking vs scoring.
- Consider alternatives that serve the same purpose with less impact.
- Document the review and the decision.
- Take legal advice if the issue is substantial.
Automated tools
If you use automated screening, ask the vendor how they test for adverse impact, what data they use and what they share. Some jurisdictions require audits. See NYC Local Law 144 and AI hiring audit checklist. Resume World shows the evidence behind each score and keeps decisions with a person, which makes a stage easier to review. We do not claim a tool can make an employer compliant. Testing and notice are the employer's duties; see compliant AI hiring and AI hiring compliance.
Making it routine
Run the calculation each quarter or after every 100 applicants for a role family, at each stage. Record the ratios and the follow-up. Over time you will see whether a change you made helped.
Common inquiries regarding this topic.
What is the four-fifths rule?
A rule of thumb from the US Uniform Guidelines on Employee Selection Procedures. If the selection rate for a group is less than four-fifths (80 percent) of the rate for the group with the highest rate, enforcement agencies will generally regard it as evidence of adverse impact.
A rule of thumb from the US Uniform Guidelines on Employee Selection Procedures. If the selection rate for a group is less than four-fifths (80 percent) of the rate for the group with the highest rate, enforcement agencies will generally regard it as evidence of adverse impact.
The Resume World Team
VerifiedProduct & hiring research, Resume World
We build the screening engine behind Resume World. Everything here comes out of working on resume parsing, scoring and hiring workflows day to day — including the parts that turned out harder than expected.
See more than just keywords.
Resume World extracts verifiable evidence from every applicant against role criteria and delivers an explained, ranked shortlist. 100% free to start with zero card required.
